Silver hyponitrite: Solubility product and complexes in aqueous ammonia

J. Inorg. Nucl. Chem.. 1961, Vol. 19, pp. 107 to 114. Pergamon Press Ltd. Printed in Northern Ireland

SILVER HYPONITRITE: SOLUBILITY PRODUCT A N D COMPLEXES IN AQUEOUS AMMONIA C. N. POLYDOROPOULOS and TH. YANNAKOPOULOS National University, Laboratory of Physical Chemistry, Solonos 104, Athens (Received 5 August 1960)

Abstract--The solubility of silver hyponitrite in aqueous ammonia has been measured over the range of concentrations 0.1-2 N NHs. The results are compatible with the assumption that the predominant complex formed in low concentrations is Ag(NHs)2+. By addition of sodium hyponitrite the solubility of silver hyponitrite in ammonia (at a constant concentration) decreases first, but for moderate and higher concentrations of Na~N~O~ it increases again, showing the formation of the complex: (NHs)AgN202-. The instability constant of the latter is found to be approximately 4 × 10-~ at 25.0°, and the solubility product of Ag~NaO~ 1.30(~0.15) × 10 1~ at the same temperature. IT has been pointed out by DIVERSm and many other investigators that silver hyponitrite is slightly soluble in concentrated solutions of hyponitrites and considerably soluble in aqueous ammonia. However nothing has been reported on the nature of the complexes formed, although dissolution of silver hyponitrite in ammonia and reprecipitation by mineral acids has been employed very often for purification purposes. In view of the interesting structure of the ion N2022- ~21and its possible mesomeric forms some attention should be paid to its co-ordination chemistry. Apart from the complex ion (('NHa)sCo)2NzO24+ studied by GRIFFITH(3) and earlier workers, there seems to be in the literature very little, if any, information on the co-ordination properties of N~O22-. We present here in brief the results of some solubility and e.m.f, measurements revealing the formation of the complex (NHa)AgN202- and we hope to be able in the future to provide some more data on the subject. The method of investigating the composition of a complex by solubility measurements under suitable conditions was introduced by BODLANDER(4) who applied it to the dissolution of silver halides in ammonia. By similar arguments one would expect the equilibria Ag2N202 ~-~ 2Ag + + N 2 0 z z Ag + + 2NH~ ~¢ Ag(NHs)2 +

(1) (2)

to be established in ammoniacal solutions saturated with Ag2N20 2, with the constants Ks

and

: (Ag+)2(N,~O22-)

K 1 : (Ag+)(N Ha)2/(Ag(N Hs) 2 ~)

(11 E. D t w R s , J. Chem. Soc. 75, 95 (1899). I'-') L. KUnN a n d E. R. LxPmrocorr, J. Amer. Chem. Soc. 78, 1820 (1956); R . J . W . LE FEVaE, W. T. OH, I. H . REECE a n d R. L. WERNER, AustralianJ. Chem. 10, 361 (1957); D . J . MILLEN, C. N. POLYDOROPOULOS a n d D. WATSON, J. Chem. Soc. 687 (1960). ~ W. P. GglFFITH, J. LEWIS a n d G . WmKINSON, J. lnorg. Nucl. Chem. 7, 38 (1958). la) G. BODLXNDEa a n d R. FITTIG, g. Ph)'s. Chem. 39, 597 (1902). 107

108

C.N. POLYDOROPOULOSand TH. YANNAKOPOULOS

respectively, where the parentheses stand for activities. Since the c o n c e n t r a t i o n of Ag + is negligible, the soluble complex can be determined analytically. I f its molality is C 0, by c o m b i n i n g the two constants a n d inserting [N2022-] ---- ½[Ag(NH3)2+] = Co/2 one o b t a i n s : 1 (2Ks~ 4 log ( f ± C 0) = ~ log \ 1(121 + -~ log (NH3) (3) where f ± is t h e m e a n a c t i v i t y coefficient of the electrolyte (Ag(NH3)2)~N20 2. Therefore a plot o f log ( f ± Co) against log (NH3) is expected to be a straight line from which Ks could be evaluated. At the same time the potential o f a silver electrode immersed into the mixture is expected to drop linearly with increasing log (NHz) according to: E = E0 +

2.3RT 3F

f±

2.3RT

log~-_ ÷ - - ~

log (2K~K1)

2 × 2.3RT ~-

log (NH3)

(4)

O n the other h a n d a d d i t i o n o f a soluble hyponitrite ( N a z N 2 0 2 ) a t c o n s t a n t a m m o n i a c o n c e n t r a t i o n would reduce the silver c o n t e n t of the solutions, as it displaces both equilibria to the left, if there were no other reactions t a k i n g place. As it will be seen, however, the results o f the m e a s u r e m e n t s can n o t be interpreted in terms o f these two reactions only. EXPERIMENTAL Sodium hyponitrite was prepared by the method of Divers-Partington (discussed elsewhere~5~) with the modification that the sodium amalgam was prepared electrolytically in a cell similar to that described by A~Ec~6~. The silver hyponitrite obtained by precipitation ofthe raw material was purified by reprecipitation from cold 0-1 N nitric acid. The solutions of sodium hyponitrite used for measurements were prepared by the action of pure silver hyponitrite (not dried) on a solution of NaI. All solubility and e.m.f, measurements were made at 25'0 :k 0"I°C. The solutions were left to establish equilibrium with some solid AgzN20~ at this temperature for about six hours. Any longer treatment might have caused serious errors due to the decomposition of hyponitrites.* Filtration (for sampling out) was effected by exerting pressure on the mixtures, being still kept in the thermostat. The solutions contained always some NaOH (0.1-0'2 N) to avoid any considerable hydrolysis of N~O22-,~ as well as to stabilize it as far as possible, c~ The silver content of the samples was determined volumetrically (KSCN) after acidification with H NO3. For the determination of the total N2022- in solution,silver nitrate was added (if necessary) in small excess, the solution was brought to a pH 5-6, and the precipitate (Ag2N~O2) washed, dissolved in 1tNO3 and titrated with KSCN. The silver electrode potentials were measured by means of the set: (--) Saturated calomel electrode ILKNOs]Isolution tested [Ag (+) immersed in the thermostat. The indicator electrode was silver-plated silver. It was standardized in the same cell repeatedly against 0.1 N AgNO3 in which the activity coefficient of Ag÷ was taken as 0.82. The potentiometer was calibrated against a certificated Weston cell. The average uncertainty of the measurements was 4-1.5 mV.~ * In a series of measurements the apparent solubility of silver hyponitrite in 0.45 M NH3 was found to change by 2-3 ~obetween 1 and 5 hr after the preparation of the mixtures. Between5 and 10 hr no change in the solubility was detectable. After 24 hr treatment a sample was found to have decomposed by 17per cent. t A higher accuracy was not attempted since, unfortunately, the,accuracy in a work on hyponitrites is anyhow limited by two factors: (a) their instability and (b) lack of sufficientliterature on even the most elementary points of the chemistry of hyponitrites, e.g. their quantitative determination. 151C. N. POLYDOROPOULOS,Chim. Chronika 9.4 A, 147 (1959). c6~E. ABELand J. PROlSL, Wien. Mh. 72, 1 (1938).

Silver hyponitrite: solubility product and complexes in aqueous ammonia TABLE 1.--SOLUBILITYO F

(NH 3)

0-091 0-192 0.495 0.827 0.952 1.444

SILVER

HYPONITRITE

IN AQUEOUS

× 10-z 0-262 0-728 2.65 5.56 6.67

0-099 0.099 0.100 0.198 0.102 0.103

11-26

-]

AMMONIA

L

NaOH (molal)

Co

109

f:~ co

0.103 0110 0"140 0.281 0.202 0,272

. . . . . . . . . . .

-2

xlO ~ 0.155 0.422 1 "48 2"67 3-47 5"52

0-59 0.58 0.56 0.48 0.52 0.49

4z .

.

.

.

E 0' -- E

(mV) 456 463 486 491 502 515

.

t

_

4

I

o

tog

(sHa)

Fro. I.--Solubility of silver hyponitrite in aqueousammonia. The points shownby × havebeendeterminedlessaccurately.

RESULTS (a)

Variation of activity co@qcients with ionic strength

Increasing a m o u n t s o f K N O a were a d d e d in a series o f solutions saturated with AgzN~O 2 at c o n s t a n t 0.45 M N H 3 a n d 0.1 N N a O H . The increase o f the solubility with increasing ionic strength (up t o / ~ = 0.45) showed a d e p e n d e n c e o f the form 2A~//, - - l o g f:~ - - 1 + ~B~/#

(5)

(where A = 0-51 a n d B = 3.29 × 107) and an effective d i a m e t e r o f the ions involved o c = 4 - 0 × 10 -8. (b)

Solubility of silver hyponitrite in ammonia

This is shown by table a n d Fig. 1 where C o is the total molality o f silver containing soluble c o m p o u n d s (in equiv, o f Ag) a n d ( N H 3) is the m o l a l i t y o f free a m m o n i a i.e. the difference o f 2C 0 from the total a m m o n i a according to e q u a t i o n (2).

110

C.N.

POLYDOROPOULOS a n d TH. YANNAKOPOULOS

In Fig. 1 curve m shows log Co, whereas curve 0t shows log (]'.Co) against log (NHa). The latter seems to be in good agreement with equation (3). The activity coefficients where calculated by means of (5). The slope of the experimentally found straight line 0t is 1.30 i 0.02 as compared to 1.33 predicted by equation (3). Furthermore the value of the constant term in (3) is found to be --1.435 + 0.015. Taking K 1 = 7.86 × 10-s ( R A N D A L L (7l) o n e finds* K, = 1.53 + 0"16 × 10-19. The potential of a silver electrode in aqueous ammonia saturated with silver hyponitrite should follow equation (4).

E-Eo mV

- 500

~

o o

- 550

t -I

I

! O

- - - ' - l o g (NH3) FIo. 2 . - - S i l v e r electrode potential in aqueous a m m o n i a saturated w i t h silver hyponitrite.

Fig. 2 shows a plot of E -- E o' against log (NHa), where E o' is the potential of the same cell for (Ag +) = 1. For moderate concentrations the slope of the curve obtained is very near to 40 mV as expected, but it deviates at higher concentrations. This probably occurs because the ratio of the two activity coefficients involved in equation (4) does not remain constant for a wide region of ionic strengths (see Table 1)(f± is the activity coefficient of a univalent cation, and f_ that of a bivalent anion). Besides at higher concentrations the formation of another complex (NH3AgN202-) would have to be taken into account, as shown below. (c) Solubility o f Ag~N202 in soluble hyponitrite at constant ammonia concentration The effect of the presence of sodium hyponitrite on the solubility of silver hyponitrite is shown in Table 2. [Ag] is the total molality of silver complexes in equiv, of Ag. For (NHa) :# 0, C O is the molality of ammoniacal silver complexes, i.e. [Ag] corrected by subtraction of the solubility for the same concentration of Na~N202 and (NH3) = 0. If only reactions (1) and (2) were to be considered, CO would be the molality of Ag(NH3)e+ and f~: Co = 1/K 1 (NH3)2(Ag +) * A correction to allow for the f o r m a t i o n of A g N H s + would lie well within the limits of the experimental (Ag(NH3)2+)/(AgNHa+)(NH3) is 7.8 × l 0 s and therefore the ratio of the two complexes can n o t be smaller t h a n 7.8 X los / 0.091 = 7.1 × 102 in the region of a m m o n i a concentrations used. ~7~ M. RANOALL and J. O. HALFORO, J. Amer. Chem. Soc. 52, 178 (1930).

error since the f o r m a t i o n constant k2 :

III

Silver hyponitrite: solubility product and complexes in aqueous ammonia TABLE 2.--SOLUBILITY OF AgzN20 ~ IN Na2N202 AT CONSTANT AMMONIA CONCENTRATION* I

[NzOz-- ] _!

[Ag]

[NaOH]

Co

p

f±

f ± Co

0"75 0'77 0"74 0"70 0'65

× t0 -~ 1 "03 0"65 0"67 0-74 0"91

0"69 0"68 0"63 0.58

3'80 3"40 3"66 3"95

E 0' E (mY) -

-

(NH3) = 0 10-~ 0"055 0"094 1 "20

×

0.0311 0-0496 0.321 0.673

i

0-270 0"293 0'683 1'231

1 "96

0.364 0-443 1.66 3"28

(NH 3) = 0.31 × 10-2 1'37

0.0069 0.0138 0.0303 0.0628 0.139

0"87 0'97 1 "19 1"64

0"84 0"91 1 "06 1 "40

0'100 0"063 0.069 0.091 0'113

0.028 0-100 0.200 0.402

5'56 5"15 6'30 8"19

5"50 5"00 5"80 6"80

(NH a) - 0.88 0-198 0"28 0.092 0-34 0.193 0"73 0.378 i-50

1 "38

i

0.121 0.100 0.151 0'268 0.514

474 499-3 505'8 515"8 524.3

491 513 520 523

* Co = molalities and f ± Co = activities of silver-ammonia complexes.

O.1 £L

E

j

8

l

~ O

J (NH~)=O.

o, . . . . . .

o .......

~

o. . . . . . . . . . . .

88

..-o

(NH3)=0.31

,

0 ----o--o----~

0.1

O, 2

--'-----r

- (NH3)

0.3 - - - -

= 0

0.4

(N~O~-)

FIG. 3.--Solubility of Ag2N~O2in Na~N2O2 at constant ammonia concentration. Continuous

lines show molalities. Dotted lines show f ± C0.

That the activity o f free silver ions decreases is shown by the drift in the electrode potential. If f ± is calculated from (5)for a univalent ion of ~ = 4.0 × 10-s, then f ± Co decreases first and at higher concentrations of Na2N~O 2 it increases again, (Fig. 3). Therefore at least one more reaction apart from (I) and (2) must be considered. (d) The solubility product of silver hyponitrite To make sure that equilibrium (1) is established the silver electrode potential was measured in a series of sodium hyponitrite solutions in equilibrium with Ag~NzO 2

112

C.N. POLYDOROPOULOSandTH. YANNAKOPOULOS

(Fig. 4). A small correction was made to the molality of N202 ~- to allow for the formation of silver-hyponitrito complexes according to Table 2. E is expected to be a linear function of log (NzO22-) according to: 2.3RT

Ks

2.3RT

E = E o' + ~ 2F log f_ --

2"r log ( N 2 0 2 2 - )

with a slope of 29.5 mV, provided there is no considerable variation in f_ (the activity coefficient of N2OzZ-). The straight line shown in Fig. 4 was drawn at a slope 29.5 mV. The points of measurements do not seem to deviate considerably. For [N2022-] = 0-01, E - E o' is found to be --485.8 ~ 2 inV. Thus K,' = Ks/f_ = 3.7 ~ 0.4 × 10-19 for # ~ 0.15. If f_ = 0.30", K s would be 1.11 ± 0.12 × 10-19. E - t:"o

mV

l -450

- 500

I -2

-2.5

I -1.5

--

,og

[N, OT]

FIG. 4 . - - S i l v e r electrode potential in Na=NzO z solution in equilibrium with Ag=N=O z. DISCUSSION

To draw conclusions, as to what substances are formed when silver hyponitrite dissolves in aqueous ammonia, the most characteristic of all series of measurements are those in Fig. 3. Evidently the simplest interpretation of the curves obtained is that they represent the sum of two functions, one hyperbolic and one parabolic. The hyperbola corresponds to the solubility of Ag2N20~ in the form of Ag(NHs)2 +, the molality of which (Ca) is expected to fall with increasing N20~ ~- according to: 1 K~½ (NHa) 2 C 1 =f-~ x -~1 × (NzOz2-) ½ where f l is its act. coefficient. The parabola indicates the formation of at least one more silver complex. Its concentration seems (by Fig. 3) to be proportional to (NzO2Z-) t (for (NH3) = const.) and to (NHz) (for (NzOz 2-) = const.). These requirements are satisfied if the reaction Ag + -F NzOs z- + NHa ~- (NHa)AgN~O2-

(6)

is assumed to take place together with (1) and (2). If it is so, the molality of this complex (C2) would be: 1

Ks ~

C~ =f-22 × K'---~ × (NHa) × (NzOzZ-)½ * Calculated from (5) f o r c ¢ = 4-0 × 10 -8 a n d / ~ = 0.15.

Silver hyponitrite: solubility p r o d u c t a n d complexes in a q u e o u s a m m o n i a

113

where/(2 is the instability constant K 2 = (Ag+)(N2022-)(NHs)/(NHaAgNzOz-). Thus Co, determined volumetrically, would be the sum of the two complexes (Co = C1 + Cz) and if it is assumed that the activity coefficients of the two complex ions are very nearly equal, and if they are replaced by f+, it follows f ± CO (Ag +)

--

(NHa) 2 + (NH3)K , /(1 K2

×

1 (Ag+) 2 -

-

where the activity of ammonia has been substituted by its molality after RANDALL(8). Therefore a plot off~Co/(Ag+ ) against I/(Ag+) 2 should be a straight line from the slope of which the ratio Ks/K 2 could be estimated. This is calculated in Table 3 from the data of Table 2 and shown by Fig. 5. TABLE 3.

f ~ Co

×

E 0' -- E (mY)

0.65 0.67 0.74 0.91

× 10 " 3'80 3"40 3-66 3-95

1

(AgO 2

474 499-3 505.8 515.8 524-3

( N H 3) -- 0.31 × l0 s 1 "02 2"76 3"56 5'25 7.31

491 513 520 523

(NHs) = 0'88 × l0 s 2"04 4"79 6"31 7"08

10 --z 1.03

1

(AgO

f~- Co (Ag e )

1016 1'04 7'61 12"7 27"6 53"5

× l0 s

N 1017

•

×

0"416 2'30 3'99 4'96

1.05 1.79

2-40 3.90 6.65

10= 0'774 1 "63 2"31 2"80

The slope of the functions in Fig. 5 is 4.32 × 10-11 for (NHs) = 0.88, and 1.06 × 10-11fOr (NHz) = 0.31. Thus K2/K s is found 2.04 × 1010 and 2"93 × 10l° respectively. F r o m the intercept the value of K 1 is found 12.5 × 10-8 (for (NHs) ,- 0.88) and 10 × 10-s (for ( N H z) = 0.31). Its actual value is 7.86 × 10-8. The inconsistency is probably due to the assumption made that f l --J~z = J~. The series for ( N H z ) = 0.31 seems more reliable as it involves moderate ionic strengths and it yields a value for K1 nearer to the correct one. Therefore K., can be taken as approximately equal to 3 × 101°Ks, and this should not be in error by more than 30 per cent. It seems therefore that there can be little doubt about the existence of equilibrium (6) in solutions of silver hyponitrite in ammonia. Even without addition of sodium hyponitrite, silver hyponitrite dissolves in aqueous ammonia to form NH~AgN20 2at a considerable extent as compared to the classical Ag(NHs)2 +. Thus a calculation of the ratio Ag(NHa)2+/NHsAgN202 - (which must be very nearly equal to K2/K~K ~ ca) M. RANDALLand J. O. HALrORD,'J. Amer. Chem. Soc. 52, 192 (1930). 8

114

C . N . POLYIX)ROPOULOSand TH. YANNAKOPOULOS

(NHa)(Ag+) ~) based on the dependence of E on the concentration of N H a (Fig. 2), shows that, even in 0.2 N N H a, the concentration of Ag(NH3)~+ is only twenty times as high as that of (NH3)AgN20~-, and that their ratio decreases down to 1-2 in 2 N N H a. Thus the value K s = 1.53 × 10-19 from Fig. 1 should be considered only as an upper limit as its calculation was made with the assumption that only the complex Ag(NHa)~+ is formed. On the other hand the value 1.11 × 10-19 from Fig. 4 may be xlO ~ 3

f+C. (Ag +) 2

1

j

(NH3) =0.31 _....--0 _--.--0

0

I

1

I

2

I

3

4

-

5 x 1 0 I?

1 (Ag+)2

FIG. 5.--Illustration of the linear relationship between f ± Co/fAg +) and 1/fAg+) 2.

too low as the activity coefficient o f N2022- for /z = 0.15 was taken f_ ----0.30. Higher f _ would seem more reasonable, e.g. 0-35 which would lead to a K 8 = 1.30 × 10-19. This is actually the mean o f 1.53, 1.11 and 1.22 × 10-19 the last of which was calculated (as not shown here) from the measurements connected with the establishment of relationship (5). Thus we take as most reliable/:8 ----- 1.30 i 0.15 × 10-19 and consequently K2 ~- 4 × 10-9. A great deal of work has been devoted to the preparation and study of silver complexes with various ligands. As a rule the ligands (usually two) coordinated to a Ag + are always identical or at least very similar. Comparatively, only a few cases of detection of silver complexes with different ligands have been reported, e.g. AgNH 3 PO42-(9), AgCNH3)z(CrO4)a3- and Ag(NHa)~CrO4 -(1°), Ag(NHa)2X 2- where X - -----BrO3-, I O n , or C1-(n) etc. Thus the data presented above may also be found useful to the study of the properties of silver ion. Acknowledgement--The authors gratefully acknowledge the support of this work by the Royal Hellenic Research Foundation.

(9) H. C. SARASWAT,Proc. Indian Acad. Sci. 30 A, 329 (1949). (to) M. B. SHCrnGOL and S. M. BIRNBAUM,Zavodskaya Lab. 15, 1027 (1949) and Chem. Abstr. 44, 1362 f (1950). (11) M. B. SHCHmOL,Zh. Obshche'~Khim. 22,721 (1952); J. Gen. Chem. U.S.S.R. (Engl. transl.) 22, 787 (1952) and Chem. Abstr. 47, 3167 c (1953).